A novel flexible shuttle vehicle scheduling problem in scenic areas: Task-divided graph-based formulation and algorithm

Abstract This study focuses on the problem of flexible shuttle scheduling (FSS) which has been observed across many scenic areas worldwide. As a contrast to the popular existing shuttle networks, tourists (passengers) in the FSS problem do not have a one-to-one matching relationship with vehicles. Each vehicle may be assigned another shuttle task after transporting tourists to a scenic area, where they would remain for a while. This creates a great challenge for the optimal scheduling of vehicles. We formulate the FSS problem using a task-divided graph-based (TDGB) mathematical model. A direct mathematical model is developed as a benchmark to show the performance of the TDGB model. A two-staged re-evaluation method is designed and validated to solve the FSS problem, which employs a part of the TDGB model to obtain the solution efficiently. Numerical experiments are conducted to quantify the superiority of the TDGB mathematical model and assess the performance of the proposed methods.